Moving average filtering with deconvolution (MAD) for hidden Markov model with filtering and correlated noise

Ibrahim M. Almanjahie, Ramzan Nazim Khan, Robin K. Milne, Takeshi Nomura, Boris Martinac

Research output: Contribution to journalArticle

Abstract

Ion channel data recorded using the patch clamp technique are low-pass filtered to remove high-frequency noise. Almanjahie et al. (Eur Biophys J 44:545-556, 2015) based statistical analysis of such data on a hidden Markov model (HMM) with a moving average adjustment for the filter but without correlated noise, and used the EM algorithm for parameter estimation. In this paper, we extend their model to include correlated noise, using signal processing methods and deconvolution to pre-whiten the noise. The resulting data can be modelled as a standard HMM and parameter estimates are again obtained using the EM algorithm. We evaluate this approach using simulated data and also apply it to real data obtained from the mechanosensitive channel of large conductance (MscL) in Escherichia coli. Estimates of mean conductances are comparable to literature values. The key advantages of this method are that it is much simpler and computationally considerably more efficient than currently used HMM methods that include filtering and correlated noise.

Original languageEnglish
Pages (from-to)383-393
Number of pages11
JournalEuropean Biophysics Journal
Volume48
Issue number4
DOIs
Publication statusPublished - May 2019

Cite this

@article{83edbc1908ed449a890ccd04c1dfd737,
title = "Moving average filtering with deconvolution (MAD) for hidden Markov model with filtering and correlated noise",
abstract = "Ion channel data recorded using the patch clamp technique are low-pass filtered to remove high-frequency noise. Almanjahie et al. (Eur Biophys J 44:545-556, 2015) based statistical analysis of such data on a hidden Markov model (HMM) with a moving average adjustment for the filter but without correlated noise, and used the EM algorithm for parameter estimation. In this paper, we extend their model to include correlated noise, using signal processing methods and deconvolution to pre-whiten the noise. The resulting data can be modelled as a standard HMM and parameter estimates are again obtained using the EM algorithm. We evaluate this approach using simulated data and also apply it to real data obtained from the mechanosensitive channel of large conductance (MscL) in Escherichia coli. Estimates of mean conductances are comparable to literature values. The key advantages of this method are that it is much simpler and computationally considerably more efficient than currently used HMM methods that include filtering and correlated noise.",
keywords = "MscL, Patch clamp, Hidden Markov models, Filter approximation, Level-dependent noise, Correlated noise, Deconvolution, EM algorithm, Parameter estimation, ION-CHANNEL CURRENTS, MECHANOSENSITIVE CHANNELS, PROBABILISTIC FUNCTIONS, STATISTICAL-ANALYSIS, MOLECULAR-BASIS, PATCH-CLAMP, IDENTIFICATION, LIKELIHOOD, KINETICS, CELLS",
author = "Almanjahie, {Ibrahim M.} and Khan, {Ramzan Nazim} and Milne, {Robin K.} and Takeshi Nomura and Boris Martinac",
year = "2019",
month = "5",
doi = "10.1007/s00249-019-01368-1",
language = "English",
volume = "48",
pages = "383--393",
journal = "European Biophysics Journal",
issn = "0175-7571",
publisher = "Springer-Verlag London Ltd.",
number = "4",

}

Moving average filtering with deconvolution (MAD) for hidden Markov model with filtering and correlated noise. / Almanjahie, Ibrahim M.; Khan, Ramzan Nazim; Milne, Robin K.; Nomura, Takeshi; Martinac, Boris.

In: European Biophysics Journal, Vol. 48, No. 4, 05.2019, p. 383-393.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Moving average filtering with deconvolution (MAD) for hidden Markov model with filtering and correlated noise

AU - Almanjahie, Ibrahim M.

AU - Khan, Ramzan Nazim

AU - Milne, Robin K.

AU - Nomura, Takeshi

AU - Martinac, Boris

PY - 2019/5

Y1 - 2019/5

N2 - Ion channel data recorded using the patch clamp technique are low-pass filtered to remove high-frequency noise. Almanjahie et al. (Eur Biophys J 44:545-556, 2015) based statistical analysis of such data on a hidden Markov model (HMM) with a moving average adjustment for the filter but without correlated noise, and used the EM algorithm for parameter estimation. In this paper, we extend their model to include correlated noise, using signal processing methods and deconvolution to pre-whiten the noise. The resulting data can be modelled as a standard HMM and parameter estimates are again obtained using the EM algorithm. We evaluate this approach using simulated data and also apply it to real data obtained from the mechanosensitive channel of large conductance (MscL) in Escherichia coli. Estimates of mean conductances are comparable to literature values. The key advantages of this method are that it is much simpler and computationally considerably more efficient than currently used HMM methods that include filtering and correlated noise.

AB - Ion channel data recorded using the patch clamp technique are low-pass filtered to remove high-frequency noise. Almanjahie et al. (Eur Biophys J 44:545-556, 2015) based statistical analysis of such data on a hidden Markov model (HMM) with a moving average adjustment for the filter but without correlated noise, and used the EM algorithm for parameter estimation. In this paper, we extend their model to include correlated noise, using signal processing methods and deconvolution to pre-whiten the noise. The resulting data can be modelled as a standard HMM and parameter estimates are again obtained using the EM algorithm. We evaluate this approach using simulated data and also apply it to real data obtained from the mechanosensitive channel of large conductance (MscL) in Escherichia coli. Estimates of mean conductances are comparable to literature values. The key advantages of this method are that it is much simpler and computationally considerably more efficient than currently used HMM methods that include filtering and correlated noise.

KW - MscL

KW - Patch clamp

KW - Hidden Markov models

KW - Filter approximation

KW - Level-dependent noise

KW - Correlated noise

KW - Deconvolution

KW - EM algorithm

KW - Parameter estimation

KW - ION-CHANNEL CURRENTS

KW - MECHANOSENSITIVE CHANNELS

KW - PROBABILISTIC FUNCTIONS

KW - STATISTICAL-ANALYSIS

KW - MOLECULAR-BASIS

KW - PATCH-CLAMP

KW - IDENTIFICATION

KW - LIKELIHOOD

KW - KINETICS

KW - CELLS

U2 - 10.1007/s00249-019-01368-1

DO - 10.1007/s00249-019-01368-1

M3 - Article

VL - 48

SP - 383

EP - 393

JO - European Biophysics Journal

JF - European Biophysics Journal

SN - 0175-7571

IS - 4

ER -